Spatial priorities for freshwater biodiversity conservation in light of catchment protection and connectivity in Europe

interventions? (4) Is there correspondence between the conservation priority of catchments and their current level of protection? Catchment protection and connectivity are seldom con- sidered in systematic conservation planning exercises for freshwater biodiversity, and such exercises are absent from the Conservation Evidence knowledge hub (http://www. conservationevidence.com). We sought to answer the above questions in three prioritization scenarios in relation to catchment protection, each with and without river connectivity to increase the relevance of the spatial priorities established for conservation policy and practice. Materials and methods We used catchments as planning units, which has several advantages over arbitrary systems of grid cells or hexagons [19, 33, 34]. We used the global HydroBASINS database [35] at level 8 (of 12), where geographical Europe (10 128 044 km2) is delineated into 18 816 catchments. Data on the distribution of species were obtained from the IUCN Red List of Threatened Spe- cies [36, 37]. Data were from fishes (n = 512 species), molluscs (n = 656), odonates (n = 124), and aquatic plants (n = 339). These groups represent a variety of trophic levels and dispersal types, and are important in ecosystem functioning and services [38–41]. We only used “extant” and “probably extant” occurrences of species in their native range. Three exceptions were the critically endangered molluscs Belgrandia moitessieri and B. varica, which had only “possibly extant” occurrences but were included because of their conservation status, and the critically

endangered fish Scardinius scardafa, which only has a single introduced population. The final database contained 4,493,267 occurrence records of 1631 species in 18,816 catchments. We prioritized catchments based on the conservation status, range-restriction and ende- mism of their species. For conservation status, we used the global Red List status, or, if it was not available (n = 334 species), the European status. Threatened species were those listed as critically endangered (CR), endangered (EN) or vulnerable (VU) [36]. Because CR species are faced with imminent extinction [36], we targeted 100% of their occurrences to be included in the optimized protected area network. For EN and VU species, for which extinction risks are lower [36], we set lower targets (75% of the occurrences for EN species and 50% for VU spe- cies) based on [25]. For all other species, at least two occurrences were specified as representa- tion targets. To estimate range restriction, we calculated range size as the total area of catchments in which the species occurred. We considered a species range-restricted under 20 000 km2 for fish and molluscs (taxa with low dispersal ability) and 50 000 km2 for odonates and aquatic plants (good dispersal ability) [25]. We targeted 25% of the occurrences of range-restricted species to be included in the optimized network. When a species qualified both as threatened and as range-restricted, conservation status was considered first as it required higher percent- ages of the species’ occurrences to be included in the optimized network. For non-threatened

species whose range was not restricted, we specified two separate occurrences as representa- tion targets. A species was considered endemic if its range was restricted to one Freshwater Ecoregion of the World [42]. We considered a catchment to hold unique species assemblages if the propor- tion of ecoregion-restricted fish and mollusc species was at least 5% of all fish and mollusc spe- cies [25]. We included the qualifying catchments (n = 190) a priori (‘locked in’ by Marxan terminology) into the prioritization. We used software Marxan for spatial prioritization [43]. Marxan heuristically searches a user-defined number of spatial configurations to identify the one that best meets the user- defined targets of species representation at the lowest cost possible. Marxan was run 1000 times in each scenario studied (see below). In each run, we used the area of planning units as a PLOS ONE Spatial priorities for freshwater biodiversity in Europe PLOS ONE | https://doi.org/10.1371/journal.pone.0267801 May 17, 2022 3 / 15